On 06/10/14 22:35, Jim Schaad wrote: > > >> -----Original Message----- From: Stephen Farrell >> [mailto:[email protected]] Sent: Monday, October 06, 2014 >> 2:24 PM To: Jim Schaad; 'Ted Lemon' Cc: [email protected]; >> 'Mike Jones'; 'The IESG'; [email protected]; draft- >> [email protected] Subject: Re: [jose] Stephen >> Farrell's Discuss on draft-ietf-jose-json-web-key- 33: (with >> DISCUSS and COMMENT) >> >> >> >> On 06/10/14 22:17, Jim Schaad wrote: >>> >>> >>>> -----Original Message----- From: Ted Lemon >>>> [mailto:[email protected]] Sent: Monday, October 06, 2014 >>>> 1:34 PM To: Jim Schaad Cc: Mike Jones; Stephen Farrell; The >>>> IESG; [email protected]; >>> draft- >>>> [email protected]; [email protected] Subject: >>>> Re: [jose] Stephen Farrell's Discuss on >>> draft-ietf-jose-json-web-key- >>>> 33: (with DISCUSS and COMMENT) >>>> >>>> On Oct 6, 2014, at 4:28 PM, Jim Schaad <[email protected]> >>>> wrote: >>>>> I worry that if we starting providing guidance to DNS names, >>>>> then we need to worry about the I18N implications. I don't >>>>> remember if these are both case sensitive and easy to do the >>>>> case conversion on. >>>> >>>> Isn't this a solved problem? You convert to the unicode >>>> presentation and then convert to the canonical case as defined >>>> in the unicode standard. >>> The >>>> worst case scenario is that you encounter some script where >>>> this rule >>> doesn't >>>> work, and that script is then in the position that all scripts >>>> are in now. >>> >>> It may be it is, however this makes an assumption that clients >>> are up on how to do this. I.e. that JavaScript is going to do it >>> right when I do a strlower function on a string. I don't know >>> that this is really the case. I would hope so but am unsure. >> >> So we're talking about key ids here. In most case where those would >> use DNS names, the code that creates the key id would know what its >> doing and dumber code would be presented with the key id and would >> not have to do the tolower(). >> >> So I would say its safe to add something like "When creating a key >> id, if the code doing so is aware that it is dealing with a DNS >> name, then that code should tolower() the DNS name before including >> those bytes in the key id." > > Yes, but if that is the case, then why does it need to be lower-cased > at all? If I say my key id is "JimSchaad.foobar" and that is my DNS > address why does it need to be lowercased? Jim
Because there will be cases where two different implementations with code try to create the same key id from its components and get it wrong otherwise. Not all cases, but some. S. > >> >> S. >> >> >>> >>> _______________________________________________ jose mailing >>> list [email protected] https://www.ietf.org/mailman/listinfo/jose >>> > _______________________________________________ jose mailing list [email protected] https://www.ietf.org/mailman/listinfo/jose
